1990 - Expressed Emotion And Auditory Evoked Potentials

ExpressedEmotion and Auditory Evoked Potentials Ulrich Hegerl,StefanPriebe,ChristiaaeWilderubeand Bruno Mtiller-Oerlinghausen THE exprcssedemotion (EE) of key relatives has been shown to predict the courseof illnessin psychiatricpatients.In this study,we examinedwhether therc might be physiological correlates to the EE index in nonbiotoeical kev relativesof parienLswith affectivepsychoses. High.EE and low.EEAIarives were compaFd concernidg thet slopes of the amptitude/stiEulus intensity function (auditory evoked Nr/Pr-component).We found that the slopes were clearly steeperin the caseof low-EE r€Iatives. In comparjng the sloies of alt key_relativeswith those of an age-matchedcontrol group without psyjriatricai-differences Iy ill partn€rs, we could find no differences.Ther€for€.the slope beuween high-EEaad low.EErelarivesdo nor seemro rflecr diff;rcn(esin rhe illness of partnerc. Wespeculatedwheth€. a steepslopeas well as low EE could be associated with an action-oriented, impulsivecommunicationstyle, which would prevent the developmentof an affectively tensecommunication Dattern. The expressed emotion (EE) of key rela. tives, a$essed by the Camberwell Famity Int€rvier,, (CFI), has been found to Dredict noi only the relaps€rate of schizopirrenics lFa-lloonet al. 1984: Hogarty et al. 1996; Hooley 1985; Leff and Vaushn 1980: Vaughn er aj. 1984) bub atso ihat of oatients with neumtic deDression {Leff and Vaughn 19801.major d;pression (Hooley 1986:Hooley €t al. 1986),recent onse! mania (Miklowitz et al. 1987) and biDolar affective disorders {Miktowitz et al. 1988). These studies show quite consistentlv that EE is a psychologiial variable of cltnical relevance{Falloon 1988; Koenissbers and Handlev 19861 The consiruct validity of tbe EE con.

cept has beeDass$sed by studying irs r€ lationstup to rhe inL€raction sryle in rhe family I Valoneer a.l.1983;Miklowit z er a.l 1984r and to rhe psychophysiologica.lreactions of patients (Sturgeon et al. 198,11 Tarrier et aI. f9?9, 1988; Valo[e et a].

r984).

We were interested in whether a bioloci. -be cal corr€late of the EE index coutd found in nonbiological relatives of psychotic patients. Such a correlaie er'ould turther suppod the validity oI the EE concept, might be helpful in finding more economic predictors of outcome, and might be useful as an indicator of chanees in EE levels during interventioD studie;. In several studies it has been found that

qti,r 4,cet MD. k R€se&chAssFt&r. Lsbora@ryof Clirical psychophvsioloay. Deodrbent or

p.y*.

D.r006Be.r,"rw.;t i'e.F;a;;J iiiiiiu" Eic iii-r. -,.!1,IT: !:ly,.lin/^or B{rrn.Eschlnale6r.senior ResistrdDpp't menrorsocilrpsychrac.y, Fr& u^iJi!',,"iifti,1r"etrf"",;#L.psyanobcls. _ _Chd,ndne WiACtub. MD, is R€*&ch Feud, DepartEeni of psychiarry. Fe UnireBiiy ot Berjl!, Berlin l\4€stl. FRC Bano MLlle-Oc inChdLs"L MD. i6 pmfessoroI prr&Dscology dd Tt)&cologl. &d Head of Laboru@

E":,1;.u,rfr:yfH"J"d"-"corocy l0E

dd [email protected];i orpsyctus rry."FRu"i".riiil, oi e*ri".

PSYCHIATRY, Vol. 53,M.y 1990

EXPRESSED EMOIION AND AI'DI'TPRY EVOKED POTENTIALS

tbe amplitude/stimulus int€nsityfunction all casesconducted by the same interview' (ASF)of s€nsoryevokedpot€ntialsis sen- er (C.W.) and rated by the same rat€r sitive to personalitytmits, suchas sensa- (S.P). The rater was trained and regardtion seeking {Buchsbaumand Pfeffer. ed as sufficiently reliable in EE ratin& baum 1971:Comoet al. 1984;Courseyet Moreover the rater was blind toward al. 1975;HaI et al. 1970;Heg€rl et al. the clinical features of the patients and 1989;Knorring 1981;Lukas 1987:Lukas was in no way involved in treatment ad' and Siecel197?;Mulins andLukas 1984; ministmtion. Four subjects wer€ excluded because Orlebekeet al. 1984;Stenberget a1.,in press;Zuckermanet al. 1974);impulsivity they were biological retatives {parents), lBanatt et a.l. 1987).and extmversion and two relatives refused to participat€ in (Soskisand Shagass1974;Stenberget al., the auditory evoked pot€ntia-I (AEP) rein Dress).Thes€studieshaveconsider€d cording. The AEP of the r€maining 15 key slopesof the ASF indicating amplitude relatives {9 males, 6 females; mean age: changeswith increasing stimulus intens' 49.6 yeaNi rangp: 32-TOyears)wereevaluated. One key relative was a fellow nun of ities. Wewereinterestedin whether nonbiolo- the patimt; the others werc spouses.One gical high-EE relatives differ from low-EE key relative had ta}€n benzodiazepines rclatives in their AsF-slopes of audrtory because of sleep disiurbances the day evoked potentials 1Nj/Pr-component)- before the recordrng. Apart from this, This AsF-slope shows satisfactory test/ rclatives were without any psychotropic r€test rcliability in healthy subjectsand is medication. Dividing the sample at the medran, rclanot significantly influenced by co-variagender, with < 2 critical Emarks were classi of astives or trm€ bles, such as age, fied as low-EE. ard relatives with > 2 crit_ sessment(Hegerlet a.l.1988). Westudied the relationship betweenthe ical remarks as high-EE. AU relatives AsF.slope and the EE index in relatives scoring >0 on the hostility scaleor scorof remitt€d outpatients with aJfectivedis- ing > 3 on the overinvolvement sca.lewere orders. Since poteDtial slope differcnces within the high-EE group. Patients living with high-EE relatives. betw€en hish-EE and low.EE relatives misht be due to differencesin the illness accoding to this definition, had a sigtifi' of iheir partners, we compar€dthe slope cantly poorer course of illness, Prcspecmeasues of the key relativeswith thoseof tively this difference was gieat€r than ag€-comparablehea.lthy controls without retrospectively, which confirms the pre_ psychiatdcally ill partners. If the illness dictive va-Iidity of EE assessment in this of the partners wereto influenc€the slope sanole (Priebeet al. 1989). For the agecomparable control group measures,then differencesbetweenthese groups we had 1? healthy drug-free volunteels two should be expected. Becausethe cortical evokedpot€Dtials (ll males, 6 females;mead age:48 yearsi are partly gpnetically determined (Buch- rang€: 32-63 yearsl without psychLiatricalsbaum19?4),weincludedonly nonbiologi- ly ill partners. These subjects were well kno*.n to the investigators and had been cal key relativ$. METHOD

Key relatives of 21 patients with bipolar affective and schizoaffective psychoses v.er€examinedwith the CFI. Patients had been on prcphylactic lithium medrcation for at least 3 years and werewithout psychotic symptoms at the time of the interview (Priebeet a.l.1989).The CFI was in PSYCHIAIRY vol.53,May 1900

partly recruited from hospital personnel. In order to rule out eny serious de{icits in auditory acuity sensation, thresholds to click stimuli wer€ monitored prior to AEP-recoding for both groups. AEP_rc_ cording took place on the same day as the CFI. The subjects were seated, eyes open, in a comfortable reclining chair. A black disc on th€ wall 3 met€rs in front of the sub' jects was present€d as a point of oienta_

HEGERL,PRIEBE,WII,DCNUBEAND MOLLER.OERLINCIIAUSEN

tion although no strict flxation was re. quested. Binaural clicks 158,68,78,88 dBHL; stimulus duration 0.9 ms; ISI 2.1 sec) were presenled in random order through headphones. Stimulus presentation, data collection, and averagirng werc controlled by a pathfinder II (Nicolet). Bandpass filters (2-pole Butterworth filters with 12 dB/octave role-offl were set at 1 and 30 Hz. Sampling was taken at a rate of 1,280 Hz with a 100 ms prestimulus and 300 ms poststimulus period. We recorded with gold'plated cup electrcdes Irom C", C,, and Ca, with linked mastoid electrodes as reference- A further electrcde was placed 1 cm above the outer corner of the l€It eye. {.ith linked mast id electrodes as reference, The responses to the first 30 clicks v{erc excluded in order to avoid habituation effects. To suppress artilact influences, a1l trials in which the voltage exceeded+50 uV in any of the 4 leads during the averaging period were automatically excluded.About 80 responses werc aveiaged at each intsnsity level. FPZ was used as gxound. Tlough-topeak am_ plitudes Nr/P, wer€ measurcd with a cur_ sor spot progmm on a monitor. The peak N, was det€rmired as the most n€gativ€ amplitude value in the period 55-140 ms, and th€ peak P, as the most positive am-

plitude value within the period 100-235 ms. The averag€dAEPS at the 4 stimulus intensities (C") are sbown lor a high-EE and a low-EE relative in Figure 1. A measure of the €te of increase of the alnpli_ tude Nr/P, with incrcasing stimulus intensity was obtained by adjusting a straight line to the amplitude values obtained at each intensity level using the least-square technique. The slope of the line indicates the amplitude change with increasing stimulus intensity. High.EE and low-EE subjects, and key relatives and controls were compared with reference to their slope values by an€lysis "group" and of variance with th€ lactor "lead" (C", Ca, Cr) as r€peated the factor measuremeDts. Additiona-Ily, the groups wer€ comparcd with r€Ierence to their N1 Pr.amplitudes using a 3-way-analysis of variance with the additional factor "intensity." RESULTS Comparing the AEPS oI the high'EE rclatives (mean age:54.3 years; rangB:32?0 yea$; 3 males, 4 fema.les)with those oI the low-EE rclatives (mean age:45.5 years; rangp: 40-54 years; 6 males, 2 fe_ males), w€ observed a st€ep€r slope of the

FiA@ 1 (0 Auditory aoked potenhials (C,) to fou stihulus int€rsiries in two €latives The low-EE subjec TU the high'EIl the of the NlPtcoDponent incrcAe dllitud€ c.itica.l rebaks) shows a boeprcnouced subiri S.M. (9 critical amdks)-

u0

PSYCHIATRI Vol. 53,M.y 1990

EXPBESSED EMOTION AND AUDITPRY EVOKED POTEI\TTIALS

ASF in the low-EE than in the high-EE relativeE(Fizure 2) The individual ASFS 1C,)of hjgh-EE a.ndlow-EE subjects ale showarin Ficure 3 Analysrs ot vaflance "8roup' (high-EE/low' with the faators '1ead' (C,. C,, C.) r€vealeda proEEI and 'group effect(Flr.r.):1a, 26;p< nounced 0.005)concerningthe ASF slope A sig:rulicant 'lead effect (Fr,6: 7.13' P<0 0051 was also lound, whjch resulbsfrom sleeper sloDesat C, bhan at C. or Ci No sigT)jficant interaciion of the factors gmup and 'lead (F,,": 2.46,p:0.10) was found. Concernine the N,/P,-ampbtucle S'way analvsis o{ v61i3nqs with the factors ''eroup' (hiehilow EE) and the repeat€d "lead" {C,. Ca, C"l .i"u"".ement factors "int€nsitY" (58,68,?8,88 dBHl) re' and vealedan inteiaction of group and inlen sity {F,.3,| 9,68: p < 0.001ltorrespondng Lo a more DronouncedaJnpltucle rncrease with incre;sine slimulus intensity in the low'EE croup. Furthermore main effecLs werc found for the facbors group (mdr'

cating higher amplitudes in tbe low-EEthan in the high-EE group)' rntenstty (indicating an amplitude inceas€ l'jth ln_ creasins inbensihi€s),and leact {m(rlcar' ing higier amplitudes at C. than at.Caor C.). The additional rnteractron or leao with 'iDtensity" indicabesmore amplrtude increasewith increasing stimulus rntensF tv'lnat C" thar at Caor C. Figure 4 the ASFS of th€ kev relatives or':15) and the aCe_comparable-controls (N=1?) ar€ shown A 2_wayanalysls o{ variance revealedno airflercnc€sin the sloDesof these two gmups. Furthermore. no relevant drfferencesex' isb concerning rangesand standard d€via' bionstsdl of bheslopesin the key rlattves and the contmls lkey r€latives. C,: mrni_ muJn= -0.8? uv/rodB.maximum=s 69 uV/ ,".". sd=r.6?i controls Cz: minimum: lb,zzuvi,,,"", maximum=6 21 uVi,oan sd=1.531. Concernina the N,'P.amplitude. a 3' wav analvsiJ o{ variance with the factor "grluo" tiontrols, kev relativesland the re"lead and D"eatedmeasurement factprs "group' iintensitv" showed no effect eJtects (43r0,: 0.40; P:0.61) Sigruficanb -leao ano were asain found for the taclors "inten;ity" and for the interaction of "lead" with "intensity: DIScussIoN

In the presetrt examination, the ASF slopes of low-DE relatives were^sho*'nJo be clearlv steeper tban lDose ot rugn-lrD rcIatives: lt appears rather unlikely that this differencebebweenhigh-EE and low' EE r€latives is due to variations in the I€ngth or severity of the illness of their partners, because key relatrves--an-o healthv controls without a menlally Lu oartne; did not differ in ASF slopes our iindinss thus provide evid€Dcefor a psy' chophlsiologiCd correlabeof the high/low33 73 EE dichotomv. we have observedst€eper slopesof the s l r m !! s i i r € n : ' r l l d B H L I ASF IAEP, Nr/Prcomponent) in m€nic Fi€!rc 2 natients iunDublished data) and Ln hrgh AmDbtud€/slinulusi!t€nsitv tunctioD(C, of ihe seekers (Hegerl et al I988I h i s t . E El > 2 c i t i c . l e m e k s l D d i h € l o d E E i < 2 iensation found that steep AS! Other studies gmuPs oitical rndks)

PSYCHIATRY,vol. 53.MaY 1900

llt

HEGEEL.PRIEBE.WILDGRUBEAND MULLE&{ERLINGHAUSEN

: : ^ "

,1 :aatt'

H Figure 3. The individual @plituderstimulus irt€lsity funciions (ASn C.) fo. hish-EE dd low'EE subjets. The ASFS@ st€e@ri! ile Iow-EEsubiects.

l+ H

key ,e oi'r.s (.= 5) c o n r , os l i ' 7 )

: 5 c o

€€ 53 3 rh ! u 3 i r e i s ' l y l d a H L I

Ampliiude/sLjhulus intensiry tulction (Cz) for key relative dd conNrcls.

tL2

slopes a-renot only related to hjgh sensalion seekjng or manic symptoms but also to extraversion and impulsivity. Although difierences in stimulatioD techniques, am' Dlitude measurcs and moda.lities render it difficult to compare various studies in the field of evokedpotetrtials (Connoly 198?). the findings published so far seem to sug_ gest thab a st€ep ASF slope is r€lated to ;n action.ori€nt€d.imDulsive attitude. It is tempting to speculate that we will find these attitudes more ofteDin low_EEthan in hish-EE subiects. Hi;h EE might be related hoan inhibit' ed, lat€ntly aggres6ive communication slyle, leadiDg ho feelings of guili and an affectively tense atmosphere,whereaslow EE might be related to a morc op€n ancl impulsive way of dealing with int€rPersona.Iproblems, preventing the accumulatioD of interpersonal tensions. This expianation is consistedt with our impr€ssion that critical remarks in the CFI mainlv reflect a recrimitratory and emotionally tense attitude and a helpless inability to react adequately to pmblems cEat€d by

PSYCHIATRY. vol. 5:1,MaY 19{,o

EXPRESSEDEMMION AND AUDITORY EVOKED PMENTTALS

the patient'sillness.If the clearrclation- interventionstudies(Leff et al. 1982)can ship betweenEE and the ASF-slop€ can be monitored by a! increaseof the ASF be replicated, it would be of interest to slope. examine whether rcduction of EE within

REFERENCES !v S., FAUTK,D. M.. dd crirical lit€.ature. Clini.al PslcholoE! ILti.u BARRATT, E. S., PRTrcHARD, (1985)5:119-39. BRAND1 M- E. Tbe Elationship berreen imlut' sjvenesssubbaits, trair a[iety, dd visual Nroo HoorE! J. M. Expressedehotior in depression:ln teractions bet@n patients md higl- vs. low_ex_ augDenting/reducirg: A topographic analysis. press€d eDotion sp[s5. Joumat of Abtunat Personotitr and lndiuiduat Diflercnces lr987l 8: Psyclolos.J (1986)95:23?-46. 43 51. HooLE\ J. M,, OR6! J-, dd TEASDAIE,J. D. L4BUCESBAUM,M. S. Avenge €voked espon* dd els ol qpregsed €moLion ed relap* in depressed stimllus intlnsity in ideniical dd lrat€msl twins. patients. Binsh ,IoumaLof Ps)c/.tdtry (1986)148: Pny siolnsicaLP sycholos! 1191412:365-70. A. Indilidu642 41. BucHssuM, M. S., dd PFETFDRB^UM, KNonRlNc, L. voN- VisuEl eoked rcsponses dd sl diff@nG in stiiulus intensity Esponse. Pslplaielel, monl)@De ondas€ in patients suflering .hbp h! sialac! ll97 ll a:19-25. ftos alcoholisb. ID H. Beslotet. ed., Th. BiotostCoMo, P. G., SrMoNS,R., od Zuc'(liMAN, M. Psy' cat Effect of ALcohoL.P)en\m Prtss. 1981. chophysiolocrcsr bgic.r indi@s of *nsaiio! s€€kinA as a tun.tion ol stimulus intosity (absLEct). KoDNrcsBERa,H. lv, Dd HANDIDY B. Expressed emoiio!: FroE prtdictjve irdd io clidcal con. Pslcioprlstobs:) (1984)2r:589. stdcr. Anencan Joumal of Psychiatry \1986) CoNNor,Lx J. F. Combentaryi ERPS suggest the im. 143:r361-73poridce of subcolticrl mechuisos in activities of life LEFF, J., ed vAUcHN, C- The intenclio! t}?ic.lly ass@ist€d with coriical tunctions. In R. mnts ed relatives dpress€d enouon '! schizd Johrson, J. w. Rohrbaush dd B- Paasur@u, phEnia dd depressjv€ neuci,s. Bntish Jolmat of eds.,Cu@nt Trendsin Euenrrebted PotennalRe Pstcrtatry {1980)136:146-53. s.ar.n IEEG Suppl. 40).Els*ier, 1987. CouRsr\ R. D., BucsssAur,{,M. S., @d FRAx(EL. LEFF,J., XurpERS.L., BER(owlrz, R., et al. A co.tiolled lrial of soci.l inteflenLion in tn€ fmilies of B. L. PersotrElity beasures ed aoked respotses schizophrnic prtients. Bttish JoumaL of Pstchiain chhnic ilso@iacs. "roLmt ol Abnomal Pst t r J ( 1 9 8 2 r) 4 r : r 2 1 - 3 4 . choloe! 11975)A4:234 44, staLu&As, J. Il. visual doked potential auementing/F FaLooN, L R. H. Ex!ressed emorioni clftnt ducing dd personaliry: Tbe wrtd augmenier is a r8. Psycholosicd!Medici^. (1988)18:269?4, sensation seke. Perso.alil! ond IndiuiduaL Dif' FArooN. L R. H.. BorD, J. L., ud McGtLL, C. !V Fanily Care of S.hi'opnrentu. Guilford Press, /€r.r.€s (198?)8:385 95. LuMs, J. H. dd SrDcDL,J. Co.dcsl Dehdbns 1984. thar auEmenL or reduce wok€d pot€ntials in cats. HAI,!, B. A,, RAPPATFI M., HoPKINS.H. X,, md S.i.n..(r9??) 196:?3 75. GRTFFTN, R- B. Evoked Espon* dd behalior in Mrrlowrz, D. J., CoLDsrErN,M. J-, FALLooN,L R. cats. Sd"zc. (1970)l?0:998-1000. H., dd DoaNE, J. A. Ilteractional @relai€s of HEGEE, U., Prmqxo, J-. ULRICH,G-, md Mo.L.iqpressd eDotion in the feilies of schizopbEn' OERllNcrrAUsEN, B. Are audrLory doked potenics. Biti.h Jaumal of Pslchiarry \rga4) 144.442 the tuponse to [thi' tia.ls suitable for prdrcijrg 8?. uD prophylslis? A study on Lhe effecis of MIXLown, D. J.. GOi-DST,IN.M. J., NUECI{,IER. releat€d me@!EDe!i, ag€ 8ade., ed persolaliint€Dsity tur.tion ih LErN, K. H., ei dl. Tle fmily ud tb€ cous€ of ty on ihe Mplitude/siiDulus mmia- ln K, HaNwec md M. J. re@lt/onei healthy volunters. Phdtudcops!.hiatry llgaal Gordst€in, €ds., Und€.stdndins Mdjor MentaL Dis' 2 1 1 3 3 63 7 , odel Fdily Pr@essPress,1987. HxcERL, U., PR@HNo,J., ULRrcH,G., dd MULE& o.RLTNGBAUSEN, B. S€DsaiionBaLilg md audi- MrKl"owrrz, D. J-, Gor,DsfErx. M. J.. NLECEEF r,ErN, K. H., et al. Fmily factors ud the course oI iory evoked poteDtials. Biological Pslchatu (1989125:179-90. bipold affetire disofier ArchiDes of General P slc hiatry lJgaal 45:225 3 1, HocARr! G. E., ANDEnsoN,C. M., RErss,D. J. et. MuLLrNs,L. F,, Dd Lues, J. H. Audrtory aueDen_ al. FMily psych@docatioD, sooal skills rlaiDins, they att€nd lhe t€rs @ sEsatior sekers-ii dd baintan&@ cheEotheapy in the EJt€rc@ stimuu labsttactl. Pryciobto/o81 ll984) 2r:589. rreaiment of 6chizophelia. ArchiD$ of Genert! ORLEBEXE, J. F., Kox, A., dd ZEILLEM^rcR.C. w kychin i} 11186143.$3-a2. of th€ Augb@ting/Educing HooLD\ J. M. ExpEssed eDotioni A evis ldisiDhibitior) md the po

PSYCEIATRY, vol.53, Mar 1990

113

IJEGDNL, PRIEBE, WILDGRUBE AND MiJLLER.OERLINGHAUSEN cessing of auditory stimllus interuily (abstracd. Pslchaphysi.bq! lrgqll 2l: 59r. P R T E B E ,S . . w T L D G R U B E .C . , a n d M U L l , t r n ' B. Lithium prcphyldis dd ex. OERLrNcHAUsxx, pressed emoiio.. Bndsh Joumot of Pst.hiata 1 1 9 8 9 ) I 5 4 r 3 9969 SosKIs. D. A., ed SHAcAss.C. Evoked Poiertial test of auClnentirg_ftdlcing- P slchoPhlsiolas! 1I974)ll:l?5-90. G., BosEx, J., md RcBERc, J. Personari STENBERG, ty dd augmentingl.educing in visn.l dd audiic ry ewked lotentials. P€rsonalit! and Indiurdual Differcnces, i press. SruRcEos. D-, TURPN, G., KUIPERS,L. et al. Psy' of schizophenic pa' chophysiological 6poNe6 tienis Lohigh ud l@ expEssed enorion relalives: A follo*'up sLudy. Enrisn JoumaL of Ps]r'cniatry 1 1 9 8 41) , 1 5 : 6 2 - 6 9 . TaRrrER,N., VAUGEN,C.. LADER,M., et al. Bodny Eactions |o peple ud events in schizopbEnics. ArchiDesof Ge^e.al Ps!chiarry (19?9136:311-1i.

ll4

K., dd N., BaRlowcloucH. C., PoRCEDDU, TARRTER, wAars, S. The assessmenr of psychophysiolo9cal to rh€ *pressed emoiion of ihe Elativ6 €&tiviiy of schizophEnic paiienis. Bit&h Jormal af Psr citatrJr 11988)152:6r8-24. vALoNE.K.. NoRmN, J. P, GoDsrEh.'. M J, dd DoANE, J. A. P@lral 8pres$d efioLion od af fectiv€ sLyle in e adolescent sdple at .isk lor scbizophFnia specinr disotde*. JoLmat of Ab' namol PstchaloE ll9€3|, 92:399 107 vAroNE, K., OoLDSTEIN.M, J. md NoRloN, J P Pdntal €xpssed emorior md psychophysiologi' cal rtactiviLy in d adolescetu smple ai lisL ior schizophFnia specrtum disordets. Joumal of Ab' natual Psrcholaq! 1L9A4)93:444 57 K- S., JonES,S, €t al- F@i_ VAUcHr, C.8., SNYD0R, rclapse- Arcnt,€s ol ly factors in schtoplEnic Qe4efll Psychiatry \l9a4l 1111,L69-11. T- M-, md SIECEL'J' M., MURTAUCH, ZUCKERVAN. Selsaiion seeking dd co|iiol ausDenrilg'Educ' 'nC. Pslchophrsiolos! \1914117:5352 42

PSYCHLTTBY.vol. 53,Mav 1990